Radiotherapy is employed for lung and upper abdominal cancer treatments. There are clear evidences that local control and survival rates are correlated with high radiation dose level. Normal tissue sparing is also critical to prevent complications, especially in concurrent chemo- and radio- therapies. Tumor targeting is hence of significant importance, which is particularly challenging for those sites affected by respirator motions. Over the years, extensive efforts have been devoted to developing novel simulation, planning, and delivery techniques to accommodate motion. Yet, due to large motion variations between planning and treatment sessions, those novel efforts are compromised by the missing of high-quality 4D image-guidance modality. Especially in the increasingly used hyper-fractionation or stereotactic body radio-surgery, the high fractional dose makes them less forgiving to targeting error. Moreover, adaptive radiation therapy (ART) holds the potential to compensate errors in dose delivery. This is particularly important for lung and upper abdominal radiotherapy because of respiratory motion and tumor shrinkage. To accurately evaluate delivered dose for treatment adaptation calls for high-quality 4D cone beam CT (4DCBCT). Recently, 4DCBCT has been developed to provide respiratory-phase resolved images for treatment guidance. Its clinical applications are limited due to the following reasons. (1) Current reconstruction algorithms require long scan protocol. Yet, even under a 4-min scans, the image quality is still inferior to that of a standard 1-min 3D CBCT of static objects, let alone CT. (2) t is susceptible to scatter contamination. Truncation problem further degrades image quality and causes missing anatomy outside the field of view. These deteriorate dose calculations accuracy. One unique feature of radiotherapy is the availability of patient-specific prior information at treatment simulation. We believe that the aforementioned problems can be solved by incorporating the prior information, as well as employing novel reconstruction methods. The overall goal of this proposal is to develop and validate a 4DCBCT reconstruction system that retrieves high-quality 4DCBCT with combined geometry and intensity accuracies under a standard 1-min 3D CBCT scan protocol. Our method reconstructs 4DCBCT via motion vector domain, which is fundamentally different from conventional approaches via image intensity domain. The feasibility of this project has been demonstrated by our preliminary studies. Our goal will be accomplished by pursuing the following specific aims (SAs). SA1. We will develop a complete 4DCBCT reconstruction system. SA2. We will validate our system and demonstrate its clinical advantages. Upon completion, a novel 4DCBCT reconstruction system will have been developed and comprehensively tested. Its clinical advantages will have been demonstrated. Clinical introduction of such a system will lead to immense benefits to lung and upper abdominal cancer patients under IGRT and future ART.